Successful Database

In order to to implement a successful database we need the following properties:

The Waterfall design process can be used to help guide us:

Requirements: describe the user requirements. Ask the users what they require from their data and what they hope to achieve, understand the scope and vision.
Analysis: produce a conceptual data model based on the data requirements. This involves providing a detailed description of the data: high and low-level properties, structure of data, data type, data range etc. You can think of the data model being produced as the logical schema for a database. Do not confuse this with the requirements section.
Design: using the conceptual data model, this step produces the relational specification, the logical schema, tables and constrains needed to satisfy the description. The first cut design (i.e. the first design) will help understand if it meets the properties we require (Completeness, Integrity, Flexibility, Efficiency and Usability). Optimise the database (or databases) for the most common requests.
Implementation: Construct a database according to the specifications of a logical schema.
Testing: uncover errors in the design, implementation, structure and associate user and management support. Checking for validation and verification.
Maintenance: it is not uncommon for database implementations to be 15 years old so maintenance is crucial. Maintenance could be introduced due to operational maintenance, porting or requirements change.